11

Aristotle on 'Being is Said in Many Vays'*

Joan Kung

-source: Ilistar¡ oJ Pbilosaphl Qaartetl¡ l, 1 / I 98ó), 3-l 8

I. The Problem

It is a well-known Aristoteiian dicrum that "rhat which is may be so calledin many ways" (Met. 1003a33). He also says,"such things as signify theschemata of predication are said tobeper s¿. For to be signifies in the samenumber of ways as these are spoken of" (Met. I0I7a2?), and he speaks of"categories of being" (e.g. at Met. 7041b28-29). Ga¡eth B. Matthewsr hasraised the following quesrion with regard to passages such as those ji-rstci teJ:

Is A¡istotle supposing in such passages that (a) rhere are diffe¡enr sensesof "being,"2 or (b) there are different kinds of being, or (c) both?

Matthews has shown that the claims that a term I'ras different senses andthat it refers to different kinds are not rwo ways of saying rhe same thing.\ü7e should note also that his question is not the same question as whetherthe categories ra¡ge over things or senses. I shall assume we may be con-fident that Arisrotle rakes himself to be talking abo't ext¡a-linguistic anclextra-conceprual entities, and I shall say a bit more about th.is in Section IVbelow. To answer the latte¡ question, however, is not to answer rhe former.

The use of the term "senses" in Matthews' question may be anachron.istic,if we suppose Aristorle himself would so phrase it. He has r-ro n,ell-developed t1'reory of meaning and does not speak of a distinct class of i¡ten-tional entities. Moreover, tl-re Greek rctm senutin(iin Aristotle's works shouldprobably seldom be translated "means."3 Nevertheless, we shall see rharMatthews'question di¡ects us ro an issue of fundamental importance for theinterpretation of Aristorle's thought. I shall argue thar we frnd Anstotlemaking the same so¡t of moves as larer thinkers who explicitly propound¡he view that "exists" has different senses. This puts him in a fundamentally

Aristatle on'Being is Said in ManT'\Y'ays' 187

clifferent camp from those who claim instead that there is one account of

what it is to exist, though there may be different kinds of existents' A

conceptual mistake is supposed to be radically distinct from mistaking a bat

fo. u ti.d or a hermir crab for a crab and to call for different responses-

diagnosis ancl correction, not fiifthef investigation. The claim about senses

is a'much srfongef ciaim. To rfeat something as a mattef of senses or a

,,purely conceptual" problem amounts to legislating in advance as to what

.u" b. discovered about teality. It is to consider the matter removed from

substantive inquiry. This same basic philosophical stance can be formulated

in a variety of ways, and a thinker need not have a fully developed theory- or

even an explicit awareness of assumptions that play a pivotal role in his

thought. B^efore turning to Aristotle, I shall illustrate briefly how some iatef

thinÉers mark off as immune from inquiry one realm or group of claims in a

number of ways which, as we shali see, afe reminiscent of one or anothef

feature of Arisiotie's tl.rought. It is my belief that this is neither an historical

nor a philosophical accident.

II. Types, Analogies and Frameworks

Although existence is usuaily not treated as a Predicate or as-a. ProPertyoo-uduyr, something like a distinction between the senses of "exists" is

.o--o.,ly found in theories of logical types or categofies that involve

ascending hierarchies oflevels ofobiects ofthought (e.g., individuals, prop-

erties, pr"operties of properties, etc.) or linguistic expressions' For example'

-o¿..., quantifrcationui logic permits existential generalization- ove¡ _bot-

to.., ieueii.rdivicluals and over predicates but requires the use of a variable

of a different style for e"istenrial generalization over predicates.o s_tandard

formal logic does not embody f...il-ter type distinctions at the levei of fi¡st

ord.r práicutes. Nevertheless, such distinctions are not uncommon in

informal conrexts, are somerimes also treated formally,t and are tacitly

assumed in some ways of treating philosopilical issues. They are thought by

some ro be of help in showing thar certain perplexing questions or problems

are ill-conceiveJ a.td .tnn...rr^ry. For example: How can I understand

anotLrer's morives or moods since I do not clireccly experience them?

Motives and moods a¡e.misconceived if they afe thought to be experiences.

They are not experiences, hence could not be "among tl¡e direct intimations

of cánsciousnesi" and I discover those of others in the same way I discove r my

o*n.6 Gilbert Ryle is a brilliant proponent of such distinctions with his

<Ioctrine of category misrakes. Ryle claims that "'existence' is not a generic

word, like '.oloi..á' or 'sexed"' but has diffe¡ent senses, so that it is absurd

to say in rhe same "logical tone of voice" that "there exist prime numbers ontülednesdays" (p. 123). It may be "perfectly Proper" to say "in o-ne logical

tone of ,roi.." ,hu, Prime numbers exist and in another that \Tednesdays

188 Aristotle: Critical Assessments

exist, but it "makes no sense" to conioin or disjoin these items. One couldbring this out by adding appropriare qualifrers, e.g. "exisrs lquantitll" and"exists lwbenJ" Bert¡and Russell, F¡ed Sommers and others also deveioptheories of types rhat have the effect of rendering some claims meaninglessor nonsensical, e.g. "An itch is a mood," "Fence is a philosophe¡," "1111 iseven prime," "Vacation is smudged," "Crime is drowsy," "His anger wastriangular," "\White is human."

Some philosophers have been less eager to embrace an explicit rrue-false-nonsense trichotomy. They hold that we can make sense of many ltrinta facienonsensical claims by reading them in a way different from that in whichtheir grammatical form would indicare rhey should be read. \firtgenstein isnot prepared to say "\(/ednesday was fat and Tüesday lean" is alu,ays non-sensical. One might, for example, have associations from one's childl-roodwhich rendered it a meaningful claim. Neve¡tl-reless, the words "fat" and"lean" cannor have their primary sense (Bedeutuzg) here, and one can usethem to say something in this sentence only if one has already mastered thatsense. Their use in this sentence could not be thei¡ primary ,¡se (Phi/. Inu.II.xi.216.) Other sentences, which are nor apparently nonsensical, would be,if read as theirgrammarical form suggesrs they should be. "I am in pain" isnot nonsense but is also not on a par with "I am epileptic." Despite itsgrammatical fo¡m, it is ciose¡ to "Owl" (Phil. lnu. 1.244.) To proceed as if a

claim cannot be taken as saying what it literally appears to say is moresubtle than to label it nonsense, but the ts¡o are nonerheless akin.

Thomas Aquinas and othe¡s employ a similar srraregy in developingtheories of analogy in which it is assumed that we have some special, suigeneris knowledge that such and such items can only be analogous, e.g.knowledge that the operations of a machine, no mamer how sophisticated,can at best be analogous to human thinking. The effect, again, is to shieldsome questions from the inspecrion and investigarion to which the resr areopen.

Some philosopi"rers are t¡oubled by the fact rhat general claims thatseemingiy must be legitimate if Rylean theory is correct are ¡enderedmeaningless by that theory, claims such as "Nothing is both a subsranceand a quantity" and "There are qualities and quantities." In the face ofsucl'lworries, Carnap adopts yet anorher straregy, which consists in blocking offan area ofdiscourse fo¡ discussion ofsuch claims and rrearing rhat area as

fundamentally different from orbers.T FIe r¡akes a basic distinction berq,een"exte¡nal" and "internal" quesrions ofexistence (pp.206-21l). An exampleof the fo¡mer is: "A¡e there things and events?" Such a question is abour the"framework" within which science and rhe man in ti're street function and isnot raised within it. Inrernal questions, on rhe orher hand, are ¡aised andanswe¡ed within a science. In the framework suggesred by an allirmativeanswer to the external question just noted, internal quesrions will include"Is there a white piece of paper on my deski " "Did King A¡thur actually

Aristotle 0n'Being is Said in Many'V'ays' 189

live?" "Are unicorns and cenlaurs real or merely imaginary?" and the like(p. 207). \ü/ithin the science a claim such as "The¡e are things" will be

regarcled as "trivial and analytic" (p.2L7). Those who demand evidence for

thi existence ofthings or p¡operties or rarional numbers before accepting a

framewo¡k which includes rhem are confused. The acceptance of fe.iection ofa framework is not a marref to be dealt with by the methods used within the

sciences (pp. 2I4-21t).All tháse posirions involve supposing that there are radically different

questions or ways of knowing sllch that adoption or knowiedge of the

framework of teners within whicir a science is conducted is different "in

principle" from knowledge obtained within the science itself, regarding

i.rtaitr mattefs as pfopaedeutic to science and to be settled in advance ofor apart from scientifrc investigation, and taking these as immune to

ques;ion, revision or correction in light of developments within the sciences.

io o.,. without these commitmenrs there will be no difference "in

principle," though there wiil often be in generality, between ¡'cateSory"'

;fru*é*otk" "logicai," or "semantical" questions and questions within a

science.8It is important to realize that although conceptual matters and differ-

ences in .u,.go.y or logical rype are thought to lie outside the realm ofsubstantive investigation, they do have profound implications for ontology

and for the sciences gene¡ally. Ifsubstance and quantity are different logical

categories, we could never discovef in any science that substances are really

quantities.one ¡eason Quine's atrack on the analytic-synrhetic distinction is so

important is that it suggesrs our knowledge cannor be parcelled olrt in this

*uy b.,, is of a piece. His reluctance ro embface distinctions of logical type

or categories ii consonant with this view.e If he is right, none of .theatremp;s to cordon off areas of rholrght in the ways surveyed will be viable.

It is my own beiief that he is right. As Putnam has pointed out, we have

come r;asonably to regard as false claims which the best minds of earlier

periods could nor have conceived to be false, claims which seemed uttefiyi--.,n. to revision from any quarter. Adherents of logical grammar would

have assured us ar rhe time of Newtonian physics that it would be a con-

ceptual blunder (nonsense, a category mistake) to suggest that so much

máss could become so much energy, bur the categories of Newtonianphysics came unstuck in the face of Einsteinian physics lt is now by no

-aun, ,lonr..rsical but indeed arguable that heat tastes the same as serine to

the normal specimen of Eschiricbia cr,li "'Whether or not Quine is right, the

issue he raises can scarcely be thought to be a trivial one.

I have dwelt on rhese aspecrs of declaring something to be a matter of

senses of adopting some close relative ofrhat view because they suggest that

even if it is correcr to claim that Matthews' qllestion about Aristotle is

somewhat anachronistic as originaliy phrased, there is an important and

I9O Aristotle: Critical Assessntents

closely ¡elated set of questions thar can and ought to be ¡aised aboutAristotle's maxim:

(a) Does he trea¡ the categories of being as knowabie jn advance of tl-re

development of tl.re sciences and such that no discovery within thesciences could call for thei¡ revision? ¿r

(b) Does he put them forwa¡d as a reasonabie, but renrative, list of thehighest gene¡a or diffe¡ent predications and/o¡ of what rhere is, wl-rilebeing prepared to modify or abandon them should the development ofthe sciences so dictate? or

(c) Is his treatment confused or is it inconsistent, perhaps changing overthe course of his career?

I shall argue that A¡istotle trears the categories of being as an affirmativeanswer to (a) would suggest.il My asserrion is put forward on the basis of anabductive inference. I claim that it provides a way of explaining whyAristotle rejects some things and acceprs orhers, why he makes cemainphilosophical moves. It enables us to give a plausible reading of passageswlrich otherwise seem ptzzling or blatantly mistaken. Moreover, rhe histor-ical development of that thought in che Stoic view of lekta and rhe medievalAristotelian notion of intentiones can then be seen as a natural ourgrowth.

IIL The Kath Hauto-Kata Sumbebékos Distinction

It may seem as if there is an immediate and damaging obstacle to anyattempr to claim, as I do, that Aristotle sees rhe categories of being asrepresenting anythinél like different senses of "being" in his early wo¡ks. Iam indebted fo¡ this objection ro Dan Devereux.t' In the Posterior Ana/ytics,usually thought to be an early work, substances are said tol:e per se, whereasnon-substances arc sumbeb&ota, accidems, which are always predicated ofsubstances and somehow dependent on rhem. \ü7ith regard to rheir exist-ence, substances are primary, accidents derivative; that is, rl're predicate"existence" applies primarily and per se to subsrance, and only derivatively,ot per accidens to non-substances. Deve¡eux compares this to a case men-tioned in the Categories in Aristotle's discussion of tl're category quanriry.Tl.rere Aristotie says white is only accidentally large q'hereas large is said pers¿ of surface. Now large is said in different ways, i.e. tt is po//ac/tos legrnnenon,

insofar as it is applied ro some things per r¿ and to others per accidens,according to Arisrotle, t'ur this does not show that "Iarge" has differentsenses in these applications. Indeed it seems ir must have the same sense (see

Topics llObI6-2> for possible support). \fhere there is a contrasr betweenper se and per accidens applications of a predicate, the predicate is notambiguous. Therefore, to say substance exists per se and items in the othercategories exist kata sunzbeltlkos is rncomparible with the "caregorical

Aristotle on'Being is Saic| in Many Wals' 191

ambiguity of being," and Aristotle does nor in his early works subscribe to

such a view.Although this case looks initially l-rarmful to my claim, on closef inspec-

tion it lenás suppoft to it. According ro the explanation offered by Aristotlefor his claim about largeness in the categories, "The white is large" cannot be

understood as saying what it literally appeafs to say, for "it is not in its own

right that [white] is called a quantity." Largeness and whiteness are

ac"cidentally present together (Cat. Ja38*b8, cp. De Int. 2lal*14, Met

1 01 5 b 17-i 9) . fn An. Post . A.22 he explains that when we say "The musical

is white" we are saying that something otlser (heteron ti) than musical, here a

man, is white. In other cases such as "The log is white," the predicate or

property belongs to the subiect but belongs only contingently or coinci-

ienlaliy. \lhire happens to be present in the individual along with the

essence, which is "said of" it.The trouble with trying ro Lrndefsrand existence along the lines of other

per se or per accidens praedicata is that in the other cases we can make sense ofsuyirrg ihe predicaie rerm refers to rhe same enriry because the hooks by

*iri.t-, i, is artached, so to speak, are different in the two cases. However, in

the case of the "is" itself, Aristotle seems to have left us no hook, or we

might almost say, it is the hooks rhemselves thatare unde¡ considefation. A

.rr.Áb.. of authors have noted that Aristotle does not in generai make our

sharp distinction between an "is" ofexistence and an "is" ofpredication. Ifthisls so, it seems likely that if Aristotle is employing what we could cali

different "is"es of pfedicarion, he is also employing what we would call

different "is"es of e*istence. If the same largeness is predicated per se ofsufface and lter accidens of log, the difference is in the lr, in Aristotle's

opinion. This is in contrasr with our usual practice of employing a single

piedication relarion. Kahn, Penner and Frede have independently main-

tained that Aristotle makes use, not just of one of even two kinds ofpfedication, but in his theory ofcategories, distinguishes ren_sorts ofp_redi-

c^tio., so that "a is F" does not rePresent one logical form but ten' Frede

argues tl-rat kategoria in its technical sense in Aristotle literally means "kind

oipredication," that kinds of predication collect predicates or properties

urrá rhut rhe term "caregory" is irsecl by extension also for the latter.li Ifthenthe "is"es ofexisrence fár the categories ofbeing correspond to the "is"es of

predicarion, the¡e is at this level no single exisrence predicabie ofboth man

and white.It is no objection to my claim ro say rhar being or existence is said only

kata sam.bebek¿s 9f non-substances because they depend on theil existence

on substances whereas the converse is not the case, for we must ask how

this depenclency is to be understood. It isn't that non-substances cannot

exist afiart from substances whereas substances can exist without non-

sr]brtu^c.r-a log has to be some color or other-so the dependency must

be explained in some other way. It may be viewed as obtaining between

I92 Aristotle: Critica/ Assessnrcnts

non-sub,stances and particuiar substances or between non-substances andsubstantial kinds or forms (as well as particular substances). I have argued inan ea¡iier paper that the forme¡ is best inrerpreted as involving a distlnctionof logical type between suches and thises, so rhat case ,,..á ,,o, be con-sidered here.la The basic idea in rhe larrer case woulcr seem ro be that inorde¡ to be qualified o¡ of some amount, erc., one must first belong ro somesubstantial kind-things can only be qualified or of some amouJt, etc. invi¡tue of being substances of some kind. Being a man and being pare are noron a pat. It follows from this that at the very least non-substances(surnbeb&ctta) and substantial kinds must be diffe¡ent. If some substantialkind were a quantity, it would have ro be one of the sunzbeb&otatt and rhedistinction would coiiapse. "lwhat is Soc¡ates?" and "I{ow much of Soc¡atesis tl-rere?" are distincr questions v4-rich require different answers. so we needto ask how A¡istotle sees this disrinction. So far as I can see Aristotle viewsit in both its employments as neither a rentarive hypothesis no¡ the fruit ofscientific endeavor, but as part of the f¡amework presupposed by anyscience. The disrinction is treated with some care and detail in the An. po.rt.and Aristotle holds elsewhere that analytika a¡e to be srudied before doingscience (Met. ro01b2-J). we do not learn in science that subjects i.ruu. r,,.,-,.properties /er se and others / er accidens . we try to discover what the essenrialproperties and othe¡ necessary att¡ibu¡es of the subjects of rhe science are.we cannot be su¡e in advance of the development of the ¡erevant scienceswhether the¡e are centaurs or gods no¡ whether the sun is such as to suffe¡eclipse (An. Po.rt. B.1) but we can be su¡e that no substantiar kind can be aquality or quantity and that some things are such that they could never tellus what a substance is. The view that some rhings ur" ,.,.h thar they coulclnever be substances is also already assumed in the'foltics (Soph. Et,22i, one ofAristotle's ea¡liest rreatises. To confuse something in the category of rela-tion o¡ of quantiry with a substance is a fallacy due to ve¡bál fo¡m thatdepends in particular on "expressing in a similai manner things which arenor rhe same" (178a4*5).

The distinction as applied to the being of substantiar kinds and theiraccidents unde¡lies Aristotle's accounr of the "distinctive mark ofsubstance," which is that it is receptive of contra¡ies Cat. 5), ancl hrsunderstanding of change. Soc¡ares remains socrates as long as he is a manwhereas I're may cease to be pale while ¡emaining what he is.t6 That ¡aturalobjects change is not somerhing Aristotle thinksneeds arguing but is rathera basic assumption in the science of naru¡e (pbys. A.2). rn st'ro¡t, the distinc-tion has importanr ¡amificarions for metaphysics a¡d other sciences, butA¡istotle seems nor to have gleaned it from them.

Aristotle on'Being is Said in Many VaYs' I93

IV. Mutually Exclusive Categories of Being

A further indication that Aristotle regards "being" as if it has many senses is

to be found in his treatment of the categories of being as mutually exclusive.

Someone might hoid that it is simply a fact about the world that those

items thar can be truly predicated of individual substances fall into disjointkinds. But if one takes the position that this is a fact on a par with the fact

that papayas come in several sexes, though admittedly more general, one

can confrdently assert it only after fairly extensive investigation ofthe extra-

linguistic and extra-conceptual world. If one is prepared antecedently to

assert that the categories are mutually exclusive and to use this as a reason

for insisting that what is evidentiy one predicate or Property must be two'one is viewing it as a principle of a different ilk. In such a context "ftrll

exciusiveness of categories is tantamount to systematic [typical] ambiguityused systematically," as has been argued by Routley and Routley (p. 222)'

Aristotle insists that everything that can be predicated of substance can be

predicated in one and only one of the ways prescribed by the schemata ofpredication.lT He claims that none of the "things spoken of according to

different schemata of predication" such as qualities, quantities, etc. "wiilanalyze either into one another or into some one thing" (Met. 1024b72_16with which see 1Ol7a22-31; 1070b1-4). N7hat happens then when we runacfoss some tefm that apPears to span more than one categofy of being and

that is at least grammatically predicated in more than one way of substance?

To the uninitiated mind the¡e would seem to be two possible moves to

make in the face of sucir an occurrence. One might (1) wonde¡ whether he

had got the categories wrong-what had seemed to be two is really one (this

would l¡e treating them as kinds), or (2) insist that the term must be

ambiguous despite appearances to the contrary. Aristotle recognizes thatthere are predicates which appear to span mo¡e than one cateSory and his

cl-raracteristic response in such cases is not to suggest that the list is incor-rect but to urge that the situation is not as it seems. He maintains eitherthat the predicate, in fact, applies to only one of the categories or that we

actually have more than one predicate, i.e., he opts for systematic ambigu-ity. It is interesting that, unlike some modern theorists, he is willing to

suppose that which category an item belongs to is not always obvious.

Nevertheless, the categories themselves are not subject to question. One

might be tempted to suppose, for example, that passive qualities and.affec-

tioás spun the cut.goties of quality and passion (being acted upon),r8 butAristotle claims at Cat.9a31_b8 tl'rat although it might appear that sweet-

ness, hotness, etc. are both qualities and affections, they prove to be onlyqualities. He goes on to consider such things as pailor and ruddiness. Itmight seem that "The man is pale" is a perfectly straightforward sentence in

which a single attribute is being predicated in a single way of some man'

but this is not tire case. It masks what are really two different kinds of

194 Aristotle: Critical Assessnents

predications and rhings predicated. On the one hand, we may be speakirrgof a pallor of compiexion which is the resulr of the man's natural consrirLr-tion or some ion¡¡ disease. On the other, we may be noting a momenrarycondit.ion perhaps broughr abour by fear. In the first case we are answeringthe question "\7hat is he like?" and predicating a qualiry of the man; in thesecond, we are answering "How is he affected?" and predicaring an affectionof him. (See also Top.A.1).)

In such cases A¡istotle is not averse ro offering as a reason for there beingtwo or more praedtcata the claim that the categories are mutually exclusive.In Book III of the Pbysics he nores that we speak of "change" with regardto substance, quality, quantiry, and locomotion. Is "change" then aunivocal term standing for some process or prope¡ry common ro rhe rele-vant categories? He emphatically denies that there is any such commonelement:

. . . it is impossible, as we assert, to find anything comnlon to these whichis neither this nor quanrum nor quale nor any of rhe other predicates.Hence neither will motion and change have ¡eference to something overand above the things mentioned, fo¡ there is nothing over and abovethem.

(200b33-201a3)

A¡istotle does, l-rowever, a few lines later offer a general cross-categoriescharacterization of change or morion, which Iiardie and Gaye translated"the fulfillment of what exists porentially, in so far as it exists porentially, ismotion" (201al0-11). Should we take rhis to indicare rhat Arisrotle afterall thinks an unambiguous term can refe¡ ro irems in more than one cat-egory? I suggest rather thar this formuia is viewed by him as a kind ofschema to be filled in and applied analogously in the ¡elevant differentcategories. That this suggesrion is on the right track seems indicated by thegenerai discussion of potenrialiry and acruality in Met.Theta 6 where hesuggests that we should not seek a definition of those terms but shouldbe content to grasp the "analogy" 1048a36). Ir is also in line with hisclaim that even the so-called "common principle" rurn out to be so only"analogously" (see e.g. An, Post.76a37-l¡2,77a22-21 .

Aristotle's theory ofexclusive categories ofbeing probably sha¡es certainothe¡ features with a number of mode¡n theories which also employonly exch-rsive categories: it takes little or no account of "va¡iable ¡efer-ring expressions" such as "rhe item under considerarion"; it is notintended to cover what would now be called "higher-order" staremenrs suchas "Man is a species" and 'Being' is said in many \vays"; and perhaps itpresupposes limited ranges of significance (see Cat.6a26-3), An.Pr.5Ib27*28, Top.A.f1). That a theory with these features is naturallyunderstood nowadays as ranging ove¡ sensed expressions ratl'rer than things

Aristotle on'Being is Said in Many'Va1's' 19>

has been shown by Routley and Routley as follows (see especially pp'

216-217 and. 227-228). "\lhat I am now thinking of is paie" is a

signiÉcant claim fo¡ I may be thinking of a person. on the orhe¡ hand,,.S"ix feet is pale" is absurd- It is somerimes thought that modern essential-

ism will .nubl. .., to explain such absufdiries-six feet is not the kind of

tbing that can be pale. This suggests that the cateSories in. an adequate

theJry of significance will range over things, regardless of how they are

.lescribe.l. BLt s..ppose that what I am now thinking of is (identical with)

six feet! S,.ch problems can pefhaps be avoided by restricting the items said

ro be identicai ro rhose huui'tg rhe same significance couplings, which wjtl

rule out "what I am now thinking of" because some predicates that can be

joined with it to make a signifrcant senrence cannot be joined with "six

i.et,,. But whar are tl-re items we are talking about then? lt seems implaus-

ible to us to say thar they are things since we v/ant to protest that "what I

am now thinking of" is surely a possible description of six feet. Further-

more, it seems iÁplausible to us that if a category theory is about things,

the significance oacerrain claims does not remain unchanged with changes

of description.Aristotie avoids having his categories range over senses of sense expfes-

sions. He refuses to admii that "Six feet is pale" is absurd or non-significant.

However, his avoidance ¡elies in paft on a move that we have seen is closely

akin to tire position that "being; has differenc senses or that beings are of

different logical types. He refuses ro fead senrences such as "Six feet is pale"

or ,,Nrhite is lu.g.' as saying what they appear to say. Six feet literally is

neither pale nor áark, so we musr understand any sentence which appears to

aftribute paleness ro six feet not as saying what its 8¡ammatical expression

would ináicate bur as having a different "logical form." It is because six feet

belongs ro something else with which pale is also associated that six feet is

sai.l tá be pale, and ihut thing may also have the attribute of being what I

am now thinking of. Aristorle does not in general regard non-synonymous

expressions as reierring to rhe same enriries, so he is uniikely to suppose that,,wl-rot I am now thinking of' and "rhe man" (or "six feet") are alternative

descriptions of the same enriry, bur there is no reason to think the entities

,o *hi.ll they refer afe senses. (See also n. 1 1 and Matthews' article

"Acciclental Unities" cited there).

V. Being Is Not a Genus

The claim that being is nor a genus, frequently reiterated throughout tl"re

corpus, does not on i,, o*n establish that Aristotle is thinking of the

categories as types as opposed to kinds (see e'g' An' Past'921>}4, S'E'

r12;14-11, n.'i.tznA3>-36, cp. E.N'1096a2i-27. Cp. also Alexander,

in Met.I24.9_126.21ancl SE.l9a3*5, Met]}O3al} and 1086b9)' Ian

196 Aristotle: CriticalAssessments

Mueller has reminded me that there is nothing preposterous about denyingthat if some number of disjoint kinds are (in the same sense), their unionmust also exist, particularly ifone believes that kinds are not conventional

but 6xed in the nature of things, as Aristotle does. He has also noted that instandard set-theory (Zermelo-Fraenkel) one can define sequences ofdisjointsets, the union of which is not a set. A¡istotle might not have the same

philosophical mot;vation in refusing to countenance a set of al1 sets but he

could have some other reason. It is certainly not unheard offor a philosopher

to hold that there is a single meaning, sense or concePt ofbeing or existence

which appiies to everything, yet to claim also that there is no single genus

or property being, and that here are ultimate divisions of things that are.

Nevertl.reless, it is less plausible to atribute such a view to Aristotle than to

attribute it to later philosophers.Although this is often overlooked, Aristotle evidently does not view

species and genera as sets.te They are more akin to what would now be

called properties on one common conception of properties, both in being" g^ppy entities" or "unsaturated," to use Frege's term, and in terms of theiridentity conditions, according to which they are the counterparts of mean-

ings (as we would put it). The genus animal and the species man are always

predicated of sometl.ring and do not exist as individuals in their own right,are not "separable." Being risible is true of all and only the same organisms

as being a man, but they are different entities and answer different ques-

tions about those individual organisms. Thus, we shall be iess misled if we

think of Aristotle's claim tha.t there is no genus being or existence not as the

denial of the existence of a class but as the claim that there is no propertybeing or existence which is common to everythinS.

Followe¡s of Frege feel justifred in making such a claim because they holdthat existence is not a property at all. "A horse exists" can be re-stated "The

concept horse is fallen under" on a Fregean anaiysis, and it is thought tl'rat

this and related analyses eliminate the need to posit existence as a prope¡ty.

There is, however, no indication of such a strategy in A¡istotle. \üflhen he

ciaims in the Metapbysics thar- "nothing different is indicated b,y the

reduplication in wording of 'one man' and 'one man being"' (1003b27-30,

cp. I0)4aI6-18), he is not anticipating Kant but noting redundancy, as

G. E. L. Owen has noted ("Snares" op.cit., P. 78).Some phiiosophers influenced by \(/ittgenstein l.rave pointed out tbat

tl-rere may be no significant traits common to all and only the members ofwhat has traditionally been regarded as a species, such as tigers or lemons,

and a parallel point migl-rt well be made about the species of a given

traditional genus. However, such points are usually thougl-rt to be anti-Aristotelian and rightly so. \ffhile Aristotle clearly enlarged some of hrs

views on definition, beginning u'ith a conception of differentiae as differ-ences fronr and late¡ coming to see them also as differentiations, there is nr,r

indication that he gave up the idea that kinds of anything must share some

Aristotle on'Being is Said in A4anl' \{/t71t' I')1

common property (see M. Furth's fortl.rcoming book on the biology and

metaphysics).Aristotle does not pursue either of these modern routes, but one might

think the¡e could be another plausible reason for him to reject the claim

tliat tlre¡e is a common property or genus being. He seems to believe that ifthere were, there could be a unive¡sai science of being and l-re may on the

other grounds suppose there can be no such science, l'rence the¡e is no genus

and the categories can be simply the ultimate genera of what there is. Hedoes speak of them as highest getñ and he holds that "a single science is one

whose domain is a single genus" (An. Post.BJa)8 cp.74b24-26, ch. 1,

16a11,-I2). He also offers a variety of reasons for rejecting a universal

science of being, but what is importanc here is that he sometimes offers as a

reason for denying that there is a genus ofbeing and a science ofbeing, the

claim tlrat being is said in many ways. (See 8.E,72I7b25-34, cp.

8.N.1096a23-2il.He does not regard this claim tl'rat there are differentcategories ofbeing as one ofwhich we can be confrdent only after that fact

that the¡e is no science ofbeing has been independently established. Rather,

the plurality ofcategories Éluarantees there can be no sucl-r science.

It is well-known that ín rhe Metaphlslcs Aristotle licenses what he some-

times calls a "science of being qua being" and some commentato¡s have

taken him to be assuming at least in these late¡ discussions that there is a

common property being whicl.r belongs to everything that is (see e.g. Ross's

comments on Gamma 2.l0o3a33ff .). If Aristotle is supposing that there are

irreducibly diffe¡ent senses of "being," or that the categories represent dil-ferent iogical types, "being" in isolation is unclear or incomplete. lüle migl.rt

think of "being" or "is" as requiring subscripts in order to be informative.

Then the science of being will have to be the science of being."6 or beingo.^,

or beingo,"",. This is nlt to say that a science of sul¡stances, for example,

cannot aiso mention qualities and quantities, but there should b'e nc>

attempt to range qualities and quantities as subjects alongsicle Particularsubstances or to predicate in the same way of the fo¡mer what is appropri-ately predicated only of substances. Aristotle's discussion of the science ofbeing in Gamma 2 opens with the declaration "That ¡'hich is may be so

called in many \\¡ays" and aclds immediately "br:t r'"ith reference to one

thing." Aristotle then compares this to what ha¡tpens with heaith and

med.icine. Are we to suppose that he has changed his mrnd about the uniryof being and is now prepared to say that in such claims as "Beorge rs,"

"Courage is," and "Fever is," tlte "is" has a single sense and a singie referent

of which we may have one science ? His subsequent remarks ancl hrs com-

parison of being with healtl'r do not suPPort such a con jecture. He seems

rather to be making the point that unlike "bankr,,",, 'n,,"

and "bank,,"","

senses of "is" and tl-reir corresponcling referents are not altogether unrelated

but dependent in various s¡ays resPectively on "is,,,,," and on substance.

Indications that this is his view are that he holds that the fundamental

198 Aristr¡tle: Critical Assessntettts

science will be the science of substances, the philosopher will examine the

principles and causes of substances (100jb16-18) and tl.rere will be as many

parts ofpl-rilosophy as there are substances (1004a2). The study ofsubstance

will, of course, also help to illuminate what is connected to substance

(I\O4a2Jff.), but homonyms, wl'ret1'rer connec¡ed or not, are not (o¡ do notdenote) the subjects of a single science (see e.g. An' Post' 97b27-31'E.E.A.8). Moreover, according to an earlier discussion of "healthy"

(T0p.A.l>), someone wl-ro argued that rosy cheeks produce health on the

grounds that being healthy is being productive ofhealth and rosy cheeks are

healthy would be comn.ritting a fallacy of equivocation. Rosy cheeks are

healthy = a sign of health * productive of health Q06b34-36).

VI. Prohibitions Against Counting

tü(/e should expect a tl-rinker to offer parallel t¡eatments of "is," "one," "same"

and "other" (Met.IOO3b22-1004a1). Thus it is not surprising that type

clistinctions often surface in informal contexts as prohibitions against con-junction (or disjunction) of certain items or against counting an item r,f one

type with those of another. Ryle te1ls us,

It may be true that there exists a cathedral in Oxford, a three-enginedbomber, and a square number between 9 and 21 . But the naive passage to

the conclusion that there are three existents, a building, a brancl ofai¡craft, and number soon leads to t¡ouble."

It may well be the case that we shall not often or ever in daily life be makingsuch claims as "That quarter-tone and her favorite shirt make two and

procrastination is a third." (tü7e occasronally might! Imagine a frustrated

mother enumerating what mosr annoys her about her teen-age daughter,

who consistently plays one note off-pitch when ¡racticing ¿ sonata, wears a

ragged T-sl'ri¡t, and puts off doing her homewo¡k). It is less clear that we

should draw Ryle's moral f¡om that. A philc.rsopher's response to such ocld

clairns provides a good indication of whether he is committed to type

clistinctions of the sort of interest here. Some like \)L V. O. Qr'rine and M.\ü(/hite, v,'ill not respond in the manner of G. Ryle, but say that the fact thatwe do not or do not usually make such assertions is due to the dissimilarityof items, and, perhaps one should add, to the circumstances of most of our

lives. It goes no way tot,ard showing that the items "exist" in differentsenses or are ofdifferent logical types or that it is senseless to say there are

rhree items in each case. (See Quine, \Y'ord and Objecl, pp. 130, 229-2i0 and

21)). One need not be taken as thereby f-iolding either that b¡ands and

numbers are alike or that they are concrete s,lbstances."Unless Aristotle in his doctrine of categories is committed to distinctions

Aristotle on'Being is Said in Mary rVays' I99

of sense or iogical type it is hard to see wl-ry he should want to deny thatalthough quarter-tones and colors are very different from each other (.M¿¡.

10t3b32-36), beirzg and being one as appiied to each need not be different.Yet he does precisely this. In every cateélo¡y, he says, the one is "some

definite nature," diffe¡ent from the one in another category, and "in no case

is its nature just this, unity" (Met.1054a9-I1). "One" behaves in the same

way as "being" and is correspondingly different in each of the categories.'Just as the¡e is not being in addition to Qtara) the what or what like or how

much" so there is no one in addition to one quantity, one quality, etc.

(10i4a1 8-1 9, cp. De An.4l0al3-22).It is not ti-rat Aristotle never says such tl-rings as "If tl-rere is to be a

substance and quality and quantity" (Phys.I85a27-28). He does, but these

utterances occur in ftamework contexts. The phrase just quoted occurs near

the beginning of the Pbysics q'here Aristotle is discussing the elements and

principles to be assumed and employed in the science of nature and con-

sidering in particular vieu's he attributes to Parmenides and lt{elissus. He

says explicitly, "Now the question whethe¡ what is is one and unchange-able, does not belong to a discussion of nature" (191b2t). It is obviously"absurd" or "impossible" (181aiO) and its proponents are muddle-headed

quibblers reasoning invalidly (18ia4-10), Although it is not necessary to

bother doing so, ifone wishes to exPose their error,

The mos¡ appropriate way of all to begin is to point out that ti'rings are

said to be in many ways, and then ask in what way they mean that all

things are one. Do they mean that tl-rere is nothing but substance, ornothing but quantity or quality?

(785 a20-23, trans. after Charlton)

It is clear that Aristotle is speaking in tlrese Passages of the f¡amework to be

assumed by the natural scientist ancl he treats N{elissus and Parmenides in a

manner vety similar to that ir-r which later thinkers would treat those they

regard as conceptually confused. iCp. Phys.184L¡21 also with the treatment

in ,l{¿¡.Gamma 3 and 7 of those wl'ro try to deny tire Principles of Non-Contradiction or Excluded Middle).

VIl. Conclusit-,n

I have argued that Aristotle's dictum "'Being'is said in many \vays" as

applied to the categories of being is best understood as tantamollnt to"'Being'has diflerent senses" or "Beings are ofdifferent logical types," in

tirat he treats the categories ofbeing as knowable in advance ofthe devel-

opment of the sciences and not subject to revisions stemming from them.

A fu1l delense of my general claim as well as of my belieFthat Aristotle

200 Aristotle:CriticalAssessments

invokes disrinctions oflogical type in part in response to difficulties he sees

in the thought of Plato and others of his predecessors would require a

cafeful examination of how eacir category functions in the specific sciences

to which it is most relevant, but that cannot be undertaken he¡e' I have

tried in my earlier paper on the sutriect ro show that the distinction between

thises and suches is a type distinction advanced in part as a resPonse to the

Third Man Argument. 'I'hat is a more fundamental distinction and plays a

larger role in l-ris thought as it develops than the distinction between when

and having, say. However, the distinction between the 6rst category ofbeing and the others is also of some importance. For example, Aristotlethinks that ifsensibie particulars afe not conglomerates ofaccidents flittingin and out of existence but come in stable kinds, many of Plato's problems

can be overcome. In addition, a careful examination of the relevant tfeatises

would show the quality-quantit)¡ distinction being invoked in a type-

theoretic way against some of his predecessors, such as Democritus and

Pythagoras. These are the two categories most often mentioned along withsubsrance in parrial Iists.

The invocarion of t1,pe distinctions strikes me as a brilliant and original

stroke on A¡istotle's part. Some of the problems in response to which he

invented rypes are still q'ith us ancl many philosophers concinue to resort to

suclr <]istinctions. My own conviction that this is a mistake does not lead me

to denigrate the impressiveness of Aristotle's acl'rievements.22

Notes

1. Gareth B. Matthews, "senses and Kinds," The Jaurnal of Ph)luopfu, vol' 69

o912), pp. \49-1tt.2. Tliá presence of a neuter "the" befcrre "being" at IO0Ja13 may provide some

sligi.rt evidence that he is speal<ing of the word.l. V¡. ¡. Ross and others do, however, frequently translate it in this way, so the

Greekless reader should be wary. See T. FI. Irwin, "Aristotle's concept of Significa-

tion," in Language and Logo.r, ed. by M. Schoheld and M. Nussbaum (Cambridge

University Pless, 1982, pp. 241-266).,1. Thi; prohibition in modern logic avoids the infamous paraclox which Russell

unearrhed in exanrinin¡¡ Frege's original system. I owe the basic idea tlrat Aris¡otle'scategories being also rellect distinc¡rons oflogical r1'pe and much helplul discussion

to Ciary Matthen's and Terry Penner.5. See e .g. Fred Sommers, "Types and Ontology," The Philt'sophical Reúeztt, voI' 12

(1y63), pp. 321-361, Arthur Fap, "fypes and Meaninglessness," AI'ind, r'oi' 69(1960), pp 4I-J1, and. R. Routiey and V. Routlel, ''Categories-expressions or

tl'rrngs?"' Thearia vc¡]r. 35 0969), pp. 215-238.6. Gilb.¡t Ryle,Concelt of \Iind(.Lctndon: Hutchinson, 1949)'p. I15.7. Rucloll Curnnp, "Errpiricisrn, Semantics, and OntoLogy," in l[eaning and

Nercstitl (Chicago: Universitl' of Chicago Press, 1956), pp.201-221.8 \ü. V O. Quine, Varld and 0bject (Cambriclge, lv{ass.: }v'IIT Press, 1!60), p' 27)'r. On tl,. lormer sec Qurn.. lto Dusmds uf Enrpirlti:m ' ,, ¡'ou/ n Lujra/ P")'t

Arisrotle on 'Being is Said in hlany Wh1'.r' 201

of Vietu (New York: Flarper and Row, 19ó1), pp. 20-+6. On the latter see e'g' his

i..pon.. to perer GeacL in the symposium "O¡ 1ü7hat Tlrere ls," Aristafelian

SocietySupltlementarlt Valunrc XXV(1911),p. 158,and Vhrdand'0bject'.01)' cit,p'229'fO. SáÉ Hilary Putnam, "The Analvtic and Synthetrc" tn h[ind, Language and

Reality (Cambriáge University Press, 197)), especially pp.47-'48. I owe the

first á*umple here to Terry Penner and rhe second to Cl-ring Kung. Recent s-t'dies

sho- thatihe pathv,ays foi sensory t¡ansduction frorn stimulus receprion (confr,rma-

tional change in the receptor prorein molecule) to the morile ¡esponse (flasellar

motion) are identical when the organism encounters the amino acid se¡ine and wlren

it swims agrin'r a hear gradient.11. In:'Aristorle onihe Snares of Ontology" (in NetL,Essals on Plata and Aristotlc,

ecl. R. Bambrough (New York: Ifumanities Press, 1965)' pp 69-95), G. E I-

owen maintains rhar Arisrotle thinks that "is" in i¡s existential use has many

senses (p. 79), but he relies heavily on Aristotle's provision of paraphtases forva'ous ixistence claims, as in h[et. H.2 and this ¡eliance is suspect, as Matthe'snotes in "Dualism and Solecisn-r," TbePhilosofbica/ Reuiew, vol. 80 (1971), p 91'

12. see "The Primary of onsia.. Aristotle's Debt to Plato" forthcominginP/atoniclnuestigations, ed. by D. J. O'Meara (\X/ashington: Catholic University ?ress)'-

13. Micl¡ael Frácle, t'Categories in Aristotle" in Stndies in Aristotle, ed' by D J'

O'Meara (\Tashington: Catholic University Ptess, 1981)' pp. I-24 See also C'.FI'Kalrn, "Qnestions and Categories" in QuestionJ, ed' bv H. F{iz (Dordrecht: Reidel,

r918).14. "A¡istotle on Thises, Suches and the Third N{an Argument," Pbrane'ris, vol' 26

(1981) pp. 201-247).tS. ÍÁ met. Delta Aristotle does not call the items in non-substance categories

sunúebékota but instead speaks of both substance and the non-substance cateélor-

ies as being per se. He reserves the tefm "accidental beings" for accidental unitiessuch as thelusical wholes s'hich he views as different entities from the man, thougl-r

he is part of them. (See Macthews "Accidental Unities" tn Langaage and Logos, olt'

rl¡. $7e would look at the corresponding locutions as different descriptions which

might well ¡efer to the same thinél). Flos'ever both,here and in.A4rr.vl.2 he giresbaJcally rhe same accounr of the "logical grammar of accidental being as is found in

tl-te Caf. and An. Past. and he continues to hold that non-substances are clependent on

substances, tl-rongh dilferent from them. For further penetrating discussion of n¡anv

relevant issres see Sandra L. Peterson, "The lf{asker Paradox," Princeton University.lctc¡oral dissertariorl, I 969.

16. See Marc Cohen, "Individual and Essence in Aristotle's lvfetaphysics," Paideia,

vol.7 (1978).17. The¡e is an apparent and puzzling exception to this general line at cat.7laJf.

Ackrill notes the anomaly aDd considers ways of dealrng rvirh it in his notes toAristot/e's Categories and De Inlerltrettttiane (Oxfotd: Clarenclon Press, 1963).

1g. I owe notice of this exampie to Abraham Ede[, "Aristotle's categories and the

Nature of Categorial Theory," ihe Rt'iew of X\etaphltic:, vol. 29 QC)75), pp'4611He takes u l^tgely opposing view oftire nature ofAristotle's categorial theorl''

19. Fo¡ "uid.n."

i." Cat.3 and Alan Code, "Aristotle: Essence and Accrdent" in

P/:ilosolthical Grouttds of Rat)oualitl': Itttetttions' Cñteg\rie:, End:, ed. by R Grandy ancl

R. \X/arner (Oxford: Clarendon Press, 1983), section 3. There is some qLrestion as towhether we must distingrLish rwo uses of eid.o.r in Arisrorie in some works, one forspecies an ci anotbe¡ for fcx- , and wheth er su bstan tial forrn i n tl¡ e \latapb'5 i¡s is rc be

,i.-ed as an jilc{ividual, bur in an¡ c¡se it seems n'e neecl no¡ and ought not tosuppose that Aris¡otle consiclers species and genera or lorm as sets.

').0. Ph)/o.ropltical Argurnents (Oxfo¡cl: Cla¡endon Press, l9'i5), pp. 1) 16.

202 Aristotle: Critica/ Asses.¡rnents

21. See Nlotton White, Touartl Reanion )tt PLilo.toltfu (New lbrk: Atheneum,

196)), pp.62ff.22.''An ancesror of this paper was read ar the meeting of the tüTestern Division of

tire Ame¡ican philosopl"rical Ássociarion in Detroit, April, 1980. I am grate6il to my

commenrarof, Ian Mueller, and to tl're otl.rer participants for helpful remarks and

criticisms. Tú" rrppo.t of the Lrsriture for Reseirch in the Hümanities, University of\rX/isconsin-Madisán, cluring part of rl.re time this pape¡ was being written is also

acknorvledged with gratitude as are tlle commenls of Dau Devereux on a fecent

vetsion.

l2Aristotle on the Subiecthood of Form*

Herbert Granger

t'source: Oxford Stndies in Ancient Philuolthl, 1)(1995), 135 59

The subjecthood of fo¡m as a criterion of its substantiality has come mofe to

the foreground of discussion in recenr yeafs, in rl're belief that it is an

important criterion in A¡isrorle's development of his iater metaphysics. In

itsiubjecthood form is taken to be a subject ofproperties, ofthe sort tirat

belong to things (concrete, particular obiects), the first-order properties olfor e*ámple, colour, weight, size, and the like. Inthe Categories subjecthood

was at rhe foref¡ont ofAristotle's thought about substance, but the traditionof Aristotelian scholarship holds that subjectl.rood was shunted aside in tireMetaphysics, when, under the regime of hylomorphism, Aristotle-came to

think thar marrer mer besr the criterion of subjecrhood, but for other

reasons could not be substance in the primary sense.r Frede, in 'substance lr-t

Aristotle's Metaphysici, and Frede and Parzig, in their commentary on

Metaphl5is¡ Zeta,' haue fecently endeavoured to reinstare the importance of

subjecthood in Aristotle's thought, and Shields has also fecenrly.molrnted a

.oigárol,. defence of the importánce of the subjecthood of lorm.r The argu-

-..,t. of Frede and Parzig and of Shields may be considered togethef

because, clespite thei¡ diflerences,a tl-reir afgumenrs coincide in important

fespecrs, ancl Shields's argumenr should receive special attention because-he

..rgug.r in a major effort to disarm tl¡e evidence against the subjecthood ofform provided by the 'celebrated Rylean passaÉle' (as Jonatl.ran Barnes calls

iüt ol De aninza 1..4, wlricl'r prima facie provides a formidable stumbling-block for any defence of tlie sr-rblecthood of form (or soul). This paper

considers rhe attempts of Frede and Parztg and of ShielJs ro rnvest form

with the kind of subjecthood that belongs ro tbe composite substance ancl

thus ro make more significant the roie of subjecthood in Aristotle's later

metaphysical thougl-rt. They argue rl-rar fr¡rm in rts subjecthood is just as

mucli a tó6e tr, a'tl-ris', as is the composite s'l¡sta¡ce in its oq'n subject-

hood. This paper finds their efforts unconvincing, and it sugÉlests that'

althougl,r A¡istorle never formally gives up subjecrhood, ]re abandons it